questionnaires within a three-month interval. A random stratifi ed cluster sample was applied to recruit the respondents. From each of the fi ve Flemish provinces in Belgium, six schools were randomly selected. Subsequently, within each selected school three classes were selected to participate in the study. The following sampling criteria were used: educational grade (fi rst, second and third grade) and the three Belgian schooling types (general secondary education; technical or artistic training; vocational training). All pupils from the selected classes were asked for their permission to take part in the survey study. The survey procedure was explained by a researcher. The students were assured verbally that their responses were anonymous and confi dential, and that no information would be passed on to teachers, parents or fellow pupils.